Principal component analysis of persistent homology rank functions with case studies of spatial point patterns, sphere packing and colloids

Persistent homology, while ostensibly measuring changes in topology, captures multiscale geometrical information. It is a natural tool for the analysis of point patterns. In this paper we explore the statistical power of the persistent homology rank functions. For a point pattern X we construct a fi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Physica. D Jg. 334; S. 99 - 117
Hauptverfasser: Robins, Vanessa, Turner, Katharine
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.11.2016
Schlagworte:
Betti numbers
Stochastic geometry
Topological data analysis
Stochastic geometry
Topological data analysis
Betti numbers
ISSN:0167-2789, 1872-8022
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Persistent homology, while ostensibly measuring changes in topology, captures multiscale geometrical information. It is a natural tool for the analysis of point patterns. In this paper we explore the statistical power of the persistent homology rank functions. For a point pattern X we construct a filtration of spaces by taking the union of balls of radius a centred on points in X, Xa=∪x∈XB(x,a). The rank function βk(X):{(a,b)∈R2:a≤b}→R is then defined by βk(X)(a,b)=rank(ι∗:Hk(Xa)→Hk(Xb)) where ι∗ is the induced map on homology from the inclusion map on spaces. We consider the rank functions as lying in a Hilbert space and show that under reasonable conditions the rank functions from multiple simulations or experiments will lie in an affine subspace. This enables us to perform functional principal component analysis which we apply to experimental data from colloids at different effective temperatures and to sphere packings with different volume fractions. We also investigate the potential of rank functions in providing a test of complete spatial randomness of 2D point patterns using the distances to an empirically computed mean rank function of binomial point patterns in the unit square. •We use the persistent homology rank function to study spatial point patterns.•The rank function is shown to be amenable to standard statistical techniques.•We demonstrate null hypothesis testing on simulated point patterns.•We develop a principal component analysis method for experimental data.•PCA of rank functions is successfully applied to colloidal and sphere-packing data.
ISSN:0167-2789
1872-8022
DOI:10.1016/j.physd.2016.03.007